Gated Context Aggregation Network for Image Dehazing and Deraining

TL;DR

This paper introduces a novel end-to-end gated context aggregation network that effectively restores haze-free images and also applies successfully to image deraining, outperforming previous methods in quality and robustness.

Contribution

The paper proposes a new network architecture with smoothed dilation and gated feature fusion, advancing state-of-the-art results in image dehazing and deraining.

Findings

Outperforms previous state-of-the-art methods in dehazing and deraining.

Uses smoothed dilation to reduce gridding artifacts.

Achieves significant quantitative and qualitative improvements.

Abstract

Image dehazing aims to recover the uncorrupted content from a hazy image. Instead of leveraging traditional low-level or handcrafted image priors as the restoration constraints, e.g., dark channels and increased contrast, we propose an end-to-end gated context aggregation network to directly restore the final haze-free image. In this network, we adopt the latest smoothed dilation technique to help remove the gridding artifacts caused by the widely-used dilated convolution with negligible extra parameters, and leverage a gated sub-network to fuse the features from different levels. Extensive experiments demonstrate that our method can surpass previous state-of-the-art methods by a large margin both quantitatively and qualitatively. In addition, to demonstrate the generality of the proposed method, we further apply it to the image deraining task, which also achieves the state-of-the-art performance. Code has been made available at https://github.com/cddlyf/GCANet.